* 2dust.061115.tgz
	Spline fitting subroutines (spline.f90 and splint.f90) are
	replaced with a more robust version (mipcub.f90). 

* 2dust.061101.tgz
	A bug fix for mapgrid.f90 - two surface brightness maps were 
	generated instead of surface brightness and optical depth maps.

	A bug fix for mapspec.f90 - not reading the 2dust output properly 
	in the anisotropic scattering mode.

	A bug fix in mapspec.f90 - incorrect array referencing in the 
	anisotropic scattering mode.

* 2dust.060803.tgz
	A bug has been fixed in dustprep.f90.  Some arrays are not set to 
	the correct size and caused some compilers to fail.
	
	A bug has been fixed in mapspec.f90.  Local densities are not 
	properly computed upon line integration over the bottom half of 
	the projected map.  The 2dust main code has not been affected.  
	However, the resulting SED and projected maps have been affected 
	(maps more so than the SED).  This bug is the cause for the 
	asymmetric looking projected map at 90 deg inclination (edge-on).  
	This bug would lead to up to 10-20% error in surface brightness 
	in the bottom half of the map and to 5-10% error in the SED.
	
	2dust models produced after mid-2004 need to be rechecked.
	
* 2dust.050805.tgz
	A minor fix in mie.f90.  The Rooij & van der Stap criterion to 
	trancate the recursion relation was not correctly implemented.

* 2dust.041202.tgz
	Expanded RT iteration (to follow Tdust for each species 
	after a converged solution is obtained with "averaged" dust)
	is fully functional.

* 2dust.040925.tgz
        Inserted readsed.f90 in place of agnprep.f90 for the public
	release version.  
		
	A bug in computing the asymptotic parameters for the KMH and
	Harrington (grain surface) averaging case was fixed.

* 2dust.040414.tgz
        Properly disabled the AGN mode in twodust.f90.

* 2dust.040409.tgz
	Reduced the number of calculations by employing another look-up
	table in the main RT iteration.  So, it will take less computation 
	time!  (However, it will need to generate a rather large look-up 
	table of local dust densities at each line integration step, and 
	it is done just before the main RT starts.) 
 	
	Also, enabled an option to follow emission contribution from 
	each dust species involved.  This is a prelude to (near future)
	release of the quantum heating (non-linear heating of small 
	grains) enabled version.  This expansion required some modification
	in the algorithm where cross sections are computed.  So, using
	this option with your old model parameters would not generate 
	exactly the same cross sections.  The differences are minor, 
	however.
	
	In order to directly get FITS images out of the mapgrid run, 
	FITSIO routines are incorporated.

* 2dust.031212.tgz
	Minor improvements.
	
	Fixed a bug in the multiple composition layer mode (improper 
	cross sections are referenced).

* 2dust.031107.tar.gz
	Updated the carriage control using "advance='no'" to follow
	the latest f90/95 standard.

	The bulk of the AGN mode is moved to a new subroutine "agnprep.f90".
	However, this extension is not part of the public release (just
	yet) and is not included in the tar package.

* 2dust.030701.tar.gz
	An erroneous mapspec.f90 subroutine was included in the 030625
	distribution.  The bad mapspec would not properly define the
	polar grid in mapping.

	mapgrid.f90 has been updated to avoid possible "array out of
	bounds" errors.

* 2dust.030625.tar.gz
	A new option is added when dust cross sections are calculated
	in dustprep.f90.  This option allows you to average some size 
	dependent quantities (such as Q values) over the size space 
	giving weights on the surface area (based on the method described 
	in Harrington et al. 1988, MNRAS, 231, 577).  This option is 
	added as a prelude to the planned future upgrade to include 
	the treatment of transiently heated particles.  Also, some 
	routines in dustprep.f90 has been renewed.
	
	The users' manual is upgraded.  The new version explains the 
	new averaging option mentioned above and has more step by step 
	example to run the suite of 2-Dust codes.

* 2dust.030608.tar.gz
	When the optical depth along each characteristic is calculated
	before the main iteration loop is initiated, a specific out of 
	bounds location in the cross section array, KK(), is occasionally 
	accessed.  Although it has been numerically taken care of (and 
	thus it does not pose threats to the results), it was recently 
	brought to my attention that such coding can end up in a compiler 
	error when compiled with some smart compilers.  Therefore, the 
	corresponding part of the code was fixed not to access out of 
	bounds array locations at all times.
	
* 2dust.030507.tar.gz
	THCRIT was not defined in mapspec.f90 for the case when
	BFLAG is activated (i.e., there is a bicone opening in
	the dust distribution).

* 2dust.030128.tar.gz
	problems in transferring double precision real numbers
	between some subroutines when compiled in double precision.

* 2dust.021212.tar.gz
	Makefile has ignored some file dependencies.

* 2dust.021105.tar.gz
	Initial Public Release.
